大数据生态系统 修仙之道 Kafka Blog

2019-04-01 Docs Language:简体中文 & English Programing Language:Kafka Website:www.geekparkhub.com OpenSource GitHub repo size in bytes GeekDeveloper:JEEP-711 Github:github.com/geekparkhub Gitee:gitee.com/geekparkhub

🐘 Kafka Technology 修仙之道 炼虚合道 🐘

Alt text

1. Kafka 概述

1.1 消息队列

enter image description here

点对点模式 (一对一,消费者主动拉取数据,消息收到后消息清除)点对点模型通常是一个基于拉取或者轮询的消息传送模型,这种模型从队列中请求信息,而不是将消息推送到客户端.这个模型的特点是发送到队列的消息被一个且只有一个接收者接收处理,即使有多个消息监听者也是如此.

发布/订阅模式 (一对多,数据生产后,推送给所有订阅者)
发布订阅模型则是一个基于推送的消息传送模型,发布订阅模型可以有多种不同的订阅者,临时订阅者只在主动监听主题时才接收消息,而持久订阅者则监听主题的所有消息,即使当前订阅者不可用,处于离线状态.

1.2 为什么需要消息队列

解耦 : 允许你独立的扩展或修改两边的处理过程,只要确保它们遵守同样的接口约束.

冗余 : 消息队列把数据进行持久化直到它们已经被完全处理,通过这一方式规避了数据丢失风险,许多消息队列所采用的”插入-获取-删除”范式中,在把一个消息从队列中删除之前,需要你的处理系统明确的指出该消息已经被处理完毕,从而确保你的数据被安全的保存直到你使用完毕.

扩展性 : 因为消息队列解耦了你的处理过程,所以增大消息入队和处理的频率是很容易的,只要另外增加处理过程即可.

灵活性&峰值处理能力 : 在访问量剧增的情况下,应用仍然需要继续发挥作用,但是这样的突发流量并不常见,如果为以能处理这类峰值访问为标准来投入资源随时待命无疑是巨大的浪费,使用消息队列能够使关键组件顶住突发的访问压力,而不会因为突发的超负荷的请求而完全崩溃.

可恢复性 : 系统的一部分组件失效时,不会影响到整个系统,消息队列降低了进程间的耦合度,所以即使一个处理消息的进程挂掉,加入队列中的消息仍然可以在系统恢复后被处理.

顺序保证 : 在大多使用场景下,数据处理的顺序都很重要,大部分消息队列本来就是排序的,并且能保证数据会按照特定的顺序来处理,(Kafka保证一个Partition内的消息的有序性).

缓冲 : 有助于控制和优化数据流经过系统的速度,解决生产消息和消费消息的处理速度不一致的情况.

异步通信 : 很多时候用户不想也不需要立即处理消息,消息队列提供了异步处理机制,允许用户把一个消息放入队列,但并不立即处理它,想向队列中放入多少消息就放多少,然后在需要的时候再去处理它们.

1.3 Kafka 简介

在流式计算中,Kafka一般用来缓存数据,Storm通过消费Kafka数据进行计算.

Apache Kafka是一个开源消息系统,是由Scala编程语言编写,Apache软件基金会开发一个开源消息系统项目.

Kafka最初是由LinkedIn公司开发,并于2011年初开源,2012年10月从Apache Incubator毕业,该项目的目标是为处理实时数据提供一个统一 / 高通量 / 低等待平台.

Kafka是一个分布式消息队列,Kafka对消息保存时根据Topic进行归类,发送消息者称为Producer,消息接受者称为Consumer,此外Kafka集群有多个Kafka实例组成,每个实例(Server)成为Broker.

无论是Kafka集群,还是Producer和consumer都依赖于Zookeeper集群保存一些Meta信息来保证系统可用性.

1.4 Kafka 架构

enter image description here

Producer : 消息生产者,向Kafka Broker发消息的客户端.

Consumer : 消息消费者,向Kafka Broker拉取消息的客户端.

Topic : 可以理解为一个队列.

Consumer Group(CG) : 这是Kafka用来实现Topic消息广播(发给所有Consumer)单播(发给任意一个Consumer)手段,一个Topic可以有多个CG,Topic消息会复制(仅仅只是概念上的复制)所有CG,但每个Partion只会把消息发给该CG中一个Consumer.
如果需要实现广播,只要每个Consumer有一个独立CG就可以了,要实现单播只要所有Consumer在同一个CG,用CG还可以将Consumer进行自由分组而不需要多次发送消息到不同的Topic.

Broker : 一台Kafka服务器就是一个Broker,一个集群由多个Broker组成,一个Broker可以容纳多个Topic.

Partition : 为了实现扩展性,一个非常大的Topic可以分布到多个Broker(即是服务器),一个Topic可以分为多个Partition,每个Partition是一个有序队列,Partition中每条消息都会被分配一个有序Id(Offset),Kafka只保证按一个Partition中顺序将消息发给Consumer,不保证一个Topic整体(多个Partition间)顺序.

Offset : Kafka存储文件都是按照offset.kafka来命名,用Offset作为名称的好处是方便查找,例如想找位于2049位置,只要找到2048.kafka文件即可,当然the first offset就是00000000000.kafka.

2. Kafka 集群部署

2.1 环境准备

2.1.1 集群规划

Server Node Zookooper Server Kafka Server
systemhub511 🖥️ zookeeper ✅ kafka ✅
systemhub611 🖥️ zookeeper ✅ kafka ✅
systemhub711 🖥️ zookeeper ✅ kafka ✅

2.1.2 Download

2.1.3 安装 Zookeeper

2.1.3.1 解压 zookeeper
[root@systemhub511 ~]#  cd /opt/software/

[root@systemhub511 software]# ll

total 512064

-rwxrwxrwx. 1 root root  35042811 Jan 17 19:18 zookeeper-3.4.10.tar.gz

[root@systemhub511 software]# tar -zxvf zookeeper-3.4.10.tar.gz -C /opt/module/

zookeeper-3.4.10/

zookeeper-3.4.10/LICENSE.txt

zookeeper-3.4.10/lib/

zookeeper-3.4.10/lib/log4j-1.2.16.LICENSE.txt

[root@systemhub511 software]#
2.1.3.2 重命名 zookeeper
[root@systemhub511 module]# mv zookeeper-3.4.10 zookeeper

[root@systemhub511 module]# ll

total 20

drwxr-xr-x.  9 root  root  4096 Feb 24 21:55 apache-tomcat

drwxr-xr-x. 10 root  root  4096 Apr 11 17:02 flume

drwxr-xr-x. 12 10011 10011 4096 Mar  3 00:42 hadoop

drwxr-xr-x.  8 uucp    143 4096 Dec 20  2017 jdk1.8.0_162

drwxr-xr-x. 10  1001  1001 4096 Mar 23  2017 zookeeper
2.1.3.3 创建 zkData目录
[root@systemhub511 zookeeper]# mkdir zkData
2.1.3.4 重命名配置文件
[root@systemhub511 conf]# mv zoo_sample.cfg zoo.cfg

[root@systemhub511 conf]# ll

total 12

-rw-rw-r--. 1 1001 1001  535 Mar 23  2017 configuration.xsl

-rw-rw-r--. 1 1001 1001 2161 Mar 23  2017 log4j.properties

-rw-rw-r--. 1 1001 1001  922 Mar 23  2017 zoo.cfg

[root@systemhub511 conf]#
2.1.3.5 配置 zoo.cfg文件

配置数据缓存路径

[root@systemhub511 zkData]# pwd

/opt/module/zookeeper/zkData

[root@systemhub511 zookeeper]# cd conf/

[root@systemhub511 conf]# vim zoo.cfg

修改配置信息

配置参数解读 : Server.A = B:C:D
A表示 标识服务器节点ID.
B表示 标识服务器节点名称.
C表示 标识服务器与集群中Leader服务器交换信息端口.
D表示 如集群中Leader服务器宕机时,需要一个端口来重新进行选举,并选出新的Leader,而这个端口就是用来执行选举时服务器相互通信端口.

# The number of milliseconds of each tick

tickTime=2000

# The number of ticks that the initial 

# synchronization phase can take

initLimit=10

# The number of ticks that can pass between 

# sending a request and getting an acknowledgement

syncLimit=5

# the directory where the snapshot is stored.

# do not use /tmp for storage, /tmp here is just 

# example sakes.

dataDir=/opt/module/zookeeper/zkData

# the port at which the clients will connect

clientPort=2181



################### Cluster ######################

server.1=systemhub511:2888:3888

server.2=systemhub611:2888:3888

server.3=systemhub711:2888:3888
2.1.3.6 Zookeeper 集群

集群模式下需配置myid文件,此文件是在dataDir目录下,此文件数据中就是A值,Zookeeper启动时读取此文件,得到的数据与zoo.cfg中配置信息比较从而判断哪个Server.

创建myid

[root@systemhub511 zookeeper]# cd zkData/

[root@systemhub511 zkData]# touch myid

[root@systemhub511 zkData]# vim myid

根据zoo.cfg服务节点配置对应id,在当前systemhub511服务器,id如1

1

集群分发

[root@systemhub511 module]# scp -r zookeeper/ root@systemhub611:/opt/module/zookeeper/

README.txt 100% 1585     1.6KB/s   00:00 

zookeeper-3.4.10-recipes-election.jar 100%   13KB  13.4KB/s   00:00    

[root@systemhub511 module]# 

[root@systemhub511 module]# scp -r zookeeper/ root@systemhub711:/opt/module/zookeeper/

README.txt 100% 1585     1.6KB/s   00:00 

zookeeper-3.4.10-recipes-election.jar 100%   13KB  13.4KB/s   00:00    

[root@systemhub511 module]#

分别配置myid文件

[root@systemhub611 module]# cd zookeeper/zkData/

[root@systemhub611 zkData]# vim myid

2

[root@systemhub711 module]# cd zookeeper/zkData/

[root@systemhub711 zkData]# vim myid

3

启动 Zookeeper Server 集群

应事先关闭集群防火墙

Start systemhub511 Server Node

[root@systemhub511 zookeeper]# bin/zkServer.sh start

ZooKeeper JMX enabled by default

Using config: /opt/module/zookeeper/bin/../conf/zoo.cfg

Starting zookeeper ... already running as process 31221.

[root@systemhub511 zookeeper]#

Start systemhub611 Server Node

[root@systemhub611 zookeeper]# bin/zkServer.sh start

ZooKeeper JMX enabled by default

Using config: /opt/module/zookeeper/bin/../conf/zoo.cfg

Starting zookeeper ... already running as process 29605.

[root@systemhub611 zookeeper]#

Start systemhub711 Server Node

[root@systemhub711 zookeeper]# bin/zkServer.sh start

ZooKeeper JMX enabled by default

Using config: /opt/module/zookeeper/bin/../conf/zoo.cfg

Starting zookeeper ... already running as process 29650.

[root@systemhub711 zookeeper]#

查看 Zookeeper Server 集群状态

systemhub511 Server Node Info

[root@systemhub511 zookeeper]# bin/zkServer.sh status

ZooKeeper JMX enabled by default

Using config: /opt/module/zookeeper/bin/../conf/zoo.cfg

Mode: follower

[root@systemhub511 zookeeper]#

systemhub611 Server Node Info

[root@systemhub611 zookeeper]# bin/zkServer.sh status

ZooKeeper JMX enabled by default

Using config: /opt/module/zookeeper/bin/../conf/zoo.cfg

Mode: leader

[root@systemhub611 zookeeper]#

systemhub711 Server Node Info

[root@systemhub711 zookeeper]# bin/zkServer.sh status

ZooKeeper JMX enabled by default

Using config: /opt/module/zookeeper/bin/../conf/zoo.cfg

Mode: follower

[root@systemhub711 zookeeper]#

2.2 Kafka 集群部署

2.2.1 解压 kafka

[root@systemhub511 software]# tar -zxvf kafka_2.11-0.11.0.0.tgz -C /opt/module/

kafka_2.11-0.11.0.0/

kafka_2.11-0.11.0.0/LICENSE

kafka_2.11-0.11.0.0/NOTICE

kafka_2.11-0.11.0.0/bin/

[root@systemhub511 software]#

2.2.2 重命名文件名称

[root@systemhub511 module]# mv kafka_2.11-0.11.0.0 kafka

[root@systemhub511 module]# ll

drwxr-xr-x.  6 root  root  4096 Jun 23  2017 kafka

[root@systemhub511 module]#

2.2.3 创建logs目录

[root@systemhub511 module]# cd kafka/

[root@systemhub511 kafka]# mkdir logs

2.2.4 修改配置文件

[root@systemhub511 kafka]# cd config/

[root@systemhub511 config]# vim server.properties

# Licensed to the Apache Software Foundation (ASF) under one or more

# contributor license agreements.  See the NOTICE file distributed with

# this work for additional information regarding copyright ownership.

# The ASF licenses this file to You under the Apache License, Version 2.0

# (the "License"); you may not use this file except in compliance with

# the License.  You may obtain a copy of the License at

#

#    http://www.apache.org/licenses/LICENSE-2.0

#

# Unless required by applicable law or agreed to in writing, software

# distributed under the License is distributed on an "AS IS" BASIS,

# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.

# See the License for the specific language governing permissions and

# limitations under the License.



# see kafka.server.KafkaConfig for additional details and defaults



############################# Server Basics #############################



# The id of the broker. This must be set to a unique integer for each broker.

broker.id=0



# Switch to enable topic deletion or not, default value is false

delete.topic.enable=true



############################# Socket Server Settings #############################



# The address the socket server listens on. It will get the value returned from 

# java.net.InetAddress.getCanonicalHostName() if not configured.

#   FORMAT:

#     listeners = listener_name://host_name:port

#   EXAMPLE:

#     listeners = PLAINTEXT://your.host.name:9092

#listeners=PLAINTEXT://:9092



# Hostname and port the broker will advertise to producers and consumers. If not set, 

# it uses the value for "listeners" if configured.  Otherwise, it will use the value

# returned from java.net.InetAddress.getCanonicalHostName().

#advertised.listeners=PLAINTEXT://your.host.name:9092



# Maps listener names to security protocols, the default is for them to be the same. See the config documentation for more details

#listener.security.protocol.map=PLAINTEXT:PLAINTEXT,SSL:SSL,SASL_PLAINTEXT:SASL_PLAINTEXT,SASL_SSL:SASL_SSL



# The number of threads that the server uses for receiving requests from the network and sending responses to the network

num.network.threads=3



# The number of threads that the server uses for processing requests, which may include disk I/O

num.io.threads=8



# The send buffer (SO_SNDBUF) used by the socket server

socket.send.buffer.bytes=102400



# The receive buffer (SO_RCVBUF) used by the socket server

socket.receive.buffer.bytes=102400



# The maximum size of a request that the socket server will accept (protection against OOM)

socket.request.max.bytes=104857600





############################# Log Basics #############################



# A comma seperated list of directories under which to store log files

log.dirs=/opt/module/kafka/logss



# The default number of log partitions per topic. More partitions allow greater

# parallelism for consumption, but this will also result in more files across

# the brokers.

num.partitions=1



# The number of threads per data directory to be used for log recovery at startup and flushing at shutdown.

# This value is recommended to be increased for installations with data dirs located in RAID array.

num.recovery.threads.per.data.dir=1



############################# Internal Topic Settings  #############################

# The replication factor for the group metadata internal topics "__consumer_offsets" and "__transaction_state"

# For anything other than development testing, a value greater than 1 is recommended for to ensure availability such as 3.

offsets.topic.replication.factor=1

transaction.state.log.replication.factor=1

transaction.state.log.min.isr=1



############################# Log Flush Policy #############################



# Messages are immediately written to the filesystem but by default we only fsync() to sync

# the OS cache lazily. The following configurations control the flush of data to disk.

# There are a few important trade-offs here:

#    1. Durability: Unflushed data may be lost if you are not using replication.

#    2. Latency: Very large flush intervals may lead to latency spikes when the flush does occur as there will be a lot of data to flush.

#    3. Throughput: The flush is generally the most expensive operation, and a small flush interval may lead to exceessive seeks.

# The settings below allow one to configure the flush policy to flush data after a period of time or

# every N messages (or both). This can be done globally and overridden on a per-topic basis.



# The number of messages to accept before forcing a flush of data to disk

#log.flush.interval.messages=10000



# The maximum amount of time a message can sit in a log before we force a flush

#log.flush.interval.ms=1000



############################# Log Retention Policy #############################



# The following configurations control the disposal of log segments. The policy can

# be set to delete segments after a period of time, or after a given size has accumulated.

# A segment will be deleted whenever *either* of these criteria are met. Deletion always happens

# from the end of the log.



# The minimum age of a log file to be eligible for deletion due to age

log.retention.hours=168



# A size-based retention policy for logs. Segments are pruned from the log as long as the remaining

# segments don't drop below log.retention.bytes. Functions independently of log.retention.hours.

#log.retention.bytes=1073741824



# The maximum size of a log segment file. When this size is reached a new log segment will be created.

log.segment.bytes=1073741824



# The interval at which log segments are checked to see if they can be deleted according

# to the retention policies

log.retention.check.interval.ms=300000



############################# Zookeeper #############################



# Zookeeper connection string (see zookeeper docs for details).

# This is a comma separated host:port pairs, each corresponding to a zk

# server. e.g. "127.0.0.1:3000,127.0.0.1:3001,127.0.0.1:3002".

# You can also append an optional chroot string to the urls to specify the

# root directory for all kafka znodes.

zookeeper.connect=systemhub511:2181,systemhub611:2181,systemhub711:2181



# Timeout in ms for connecting to zookeeper

zookeeper.connection.timeout.ms=6000





############################# Group Coordinator Settings #############################



# The following configuration specifies the time, in milliseconds, that the GroupCoordinator will delay the initial consumer rebalance.

# The rebalance will be further delayed by the value of group.initial.rebalance.delay.ms as new members join the group, up to a maximum of max.poll.interval.ms.

# The default value for this is 3 seconds.

# We override this to 0 here as it makes for a better out-of-the-box experience for development and testing.

# However, in production environments the default value of 3 seconds is more suitable as this will help to avoid unnecessary, and potentially expensive, rebalances during application startup.

group.initial.rebalance.delay.ms=0

2.2.5 配置环境变量

[root@systemhub511 kafka]# pwd

/opt/module/kafka

[root@systemhub511 kafka]# vim /etc/profile

配置信息

## SET KAFKA_HOME

export KAFKA_HOME=/opt/module/kafka

export PATH=$PATH:$KAFKA_HOME/bin

更新配置文件

[root@systemhub511 kafka]# source /etc/profile

[root@systemhub511 kafka]# echo $KAFKA_HOME

/opt/module/kafka

2.2.6 分发 kafka集群

[root@systemhub511 module]# scp -r kafka/ root@systemhub611:/opt/module/kafka/

connect-file-source.properties 100%  881     0.9KB/s   00:00    

.server.properties.swp 100%   16KB  16.0KB/s   00:00    

connect-file-sink.properties 100%  883     0.9KB/s   00:00

[root@systemhub511 module]#

[root@systemhub511 module]# scp -r kafka/ root@systemhub711:/opt/module/kafka/

connect-file-source.properties 100%  881     0.9KB/s   00:00    

.server.properties.swp 100%   16KB  16.0KB/s   00:00 

[root@systemhub511 module]#

分别修改配置文件,broker.id不得重复

[root@systemhub611 ~]# cd /opt/module/kafka/config/

[root@systemhub611 config]# vim server.properties

# The id of the broker. This must be set to a unique integer for each broker.

broker.id=1

[root@systemhub711 ~]# cd /opt/module/kafka/config/

[root@systemhub711 config]# vim server.properties

# The id of the broker. This must be set to a unique integer for each broker.

broker.id=2

2.2.7 启动 kafka集群

因KafkaServer依赖于ZookeeperServer,所以要事先开启ZookeeperServer.

[root@systemhub511 kafka]#  bin/kafka-server-start.sh config/server.properties

[1] 23152

[root@systemhub611 kafka]#  bin/kafka-server-start.sh config/server.properties

[1] 22647

[root@systemhub711 kafka]# bin/kafka-server-start.sh config/server.properties &

[1] 22736

2.2.8 关闭 kafka集群

[root@systemhub511 kafka]# bin/kafka-server-stop.sh stop

[root@systemhub611 kafka]# bin/kafka-server-stop.sh stop

[root@systemhub711 kafka]# bin/kafka-server-stop.sh stop

2.3 Kafka 指令

2.3.1 创建Topic

参数说明 :
--topic 定义topic名称
--replication-factor 定义副本数量
--partitions 定义分区数量

[root@systemhub511 kafka]# bin/kafka-topics.sh --create --zookeeper systemhub511:2181 -partitions 2 --replication-factor 2 --topic topic001

Created topic "topic001".

[root@systemhub511 kafka]#

2.3.2 查看当前服务中所有Topic

[root@systemhub511 kafka]# bin/kafka-topics.sh --list --zookeeper systemhub511:2181

topic001

[root@systemhub511 kafka]#

2.3.3 生产者

[root@systemhub511 kafka]# bin/kafka-console-producer.sh --broker-list systemhub511:9092 --topic topic001

>hello kafka

2.3.4 消费者

过时版本语法

[root@systemhub711 kafka]# bin/kafka-console-consumer.sh --zookeeper systemhub511:2181 --topic topic001

Using the ConsoleConsumer with old consumer is deprecated and will be removed in a future major release. Consider using the new consumer by passing [bootstrap-server] instead of [zookeeper].

hello kafka

新版本语法

[root@systemhub711 kafka]# bin/kafka-console-consumer.sh --bootstrap-server systemhub511:9092 --topic topic001 --from-beginning



hello kafka

2.3.5 查看Topic详情

[root@systemhub511 kafka]# bin/kafka-topics.sh --describe --zookeeper systemhub511 --topic topic001

Topic:topic001  PartitionCount:2        ReplicationFactor:2     Configs:

        Topic: topic001 Partition: 0    Leader: 1       Replicas: 1,0   Isr: 0,1

        Topic: topic001 Partition: 1    Leader: 2       Replicas: 2,1   Isr: 1,2

[root@systemhub511 kafka]#

2.3.6 删除Topic

需要在server.properties配置文件中设置delete.topic.enable=true
否则只是标记删除或者直接重启

[root@systemhub511 kafka]# bin/kafka-topics.sh --delete --zookeeper systemhub511:2181 --topic topic001

Topic topic001 is marked for deletion.

Note: This will have no impact if delete.topic.enable is not set to true.

[root@systemhub511 kafka]#

2.4 Kafka 配置信息

2.4.1 Broker 配置信息

属性 默认值 描述
broker.id 0/1/2/3/… 必填参数,broker唯一标识.
log.dirs /tmp/kafka-logs Kafka数据存放的目录,可以指定多个目录,中间用逗号分隔,当新partition被创建时会被存放到当前存放partition最少的目录.
port 9092 BrokerServer接受客户端连接端口号
zookeeper.connect null Zookeeper连接串格式为:hostname1:port1,hostname2:port2,hostname3:port3,注意,此配置允许指定一个zookeeper路径来存放此kafka集群所有数据,为了与其他应用集群区分开,建议在此配置中指定本集群存放目录格式为:hostname1:port1,hostname2:port2,hostname3:port3/chroot/path,要注意的是,消费者参数要和此参数一致.
message.max.bytes 1000000 服务器可以接收到最大消息的大小,注意此参数要和consumer的maximum.message.size值大小一致,否则会因为生产者生产消息太大导致消费者无法消费.
num.io.threads 8 服务器用来执行读写请求IO线程数,此参数数量至少要等于服务器上磁盘数量.
queued.max.requests 500 I/O线程可以处理请求队列大小,若实际请求数超过此大小,网络线程将停止接收新请求.
socket.send.buffer.bytes 100 * 1024 The SO_SNDBUFF buffer the server prefers for socket connections
socket.receive.buffer.bytes. field2 field3
field1 100 * 1024 The SO_RCVBUFF buffer the server prefers for socket connections.
socket.request.max.bytes 100 * 1024 * 1024 服务器允许请求最大值,用来防止内存溢出,其值应该小于Java heap size.
num.partitions 1 默认partition数量,如果topic在创建时没有指定partition数量,默认使用此值,建议改为5.
log.segment.bytes 1024 * 1024 * 1024 Segment文件大小,超过此值将会自动新建一个segment,此值可以被topic级别参数覆盖.
log.roll.{ms,hours} 24 * 7 hours 新建segment文件时间,此值可以被topic级别参数覆盖.
log.retention.{ms,minutes,hours} 7 days Kafka segment log保存周期,保存周期超过此时间日志就会被删除,此参数可以被topic级别参数覆盖,数据量大时建议减小此值.
log.retention.bytes -1 每个partition最大容量,若数据量超过此值,partition数据将会被删除,注意这个参数控制每个partition而不是topic,此参数可以被log级别参数覆盖.
log.retention.check.interval.ms 5 minutes 删除策略的检查周期
auto.create.topics.enable true 自动创建topic参数,建议此值设置为false,严格控制topic管理,防止生产者错写topic.
default.replication.factor 1 默认副本数量,建议改为2
replica.lag.time.max.ms 10000 在此窗口时间内没有收到follower fetch请求,leader会将其从ISR(in-syncreplicas)中移除.
replica.lag.max.messages 4000 如果replica节点落后leader节点此值大小消息数量,leader节点就会将其从ISR中移除.
replica.socket.timeout.ms 30 * 1000 replica向leader发送请求的超时时间.
replica.socket.receive.buffer.bytes 64 * 1024 The socket receive buffer for network requests tothe leader for replicating data.
replica.fetch.max.bytes 1024 * 1024 The number of byes of messages to attempt to fetch for each partition in the fetch requests the replicas send to the leader.
replica.fetch.wait.max.ms 500 The maximum amount of time towait time for data to arrive on the leader in the fetch requests sent by the replicas to the leader.
num.replica.fetchers 1 Number of threads used to replicate messages from leaders. Increasing this value can increase the degree of I/O parallelism in thefollower broker
fetch.purgatory.purge.interval.requests 1000 The purge interval (in number of requests) of the fetch request purgatory.
zookeeper.session.timeout.ms 6000 ZooKeeper session超时时间,如果在此时间内server没有向zookeeper发送心跳,zookeeper就会认为此节点已挂掉,此值太低导致节点容易被标记死亡,若太高会导致太迟发现节点死亡.
zookeeper.connection.timeout.ms 6000 客户端连接zookeeper超时时间.
controlled.shutdown.enable true 允许broker shutdown,如果启用broker在关闭之前会把它上面所有leaders转移到其它brokers上,建议启用增加集群稳定性.
auto.leader.rebalance.enable true If this is enabled the controller will automatically try to balance leadership for partitions among the brokers by periodically returning leadership to the “preferred” replica for each partition if it is available.
leader.imbalance.per.broker.percentage 10 The percentage of leader imbalance allowed per broker. The controller will rebalance leadership if this ratio goes above the configured value per broker
delete.topic.enable false 启用deletetopic参数,建议设置为true.

2.4.2 Producer 配置信息

属性 默认值 描述
metadata.broker.list 启动时producer查询brokers列表,可以是集群中所有brokers一个子集,注意这个参数只是用来获取topic元信息,producer会从元信息中挑选合适的broker并与之建立socket连接,格式为:host1:port1,host2:port2
request.timeout.ms 10000 Broker等待ack超时时间,若等待时间超过此值,会返回客户端错误信息.
producer.type sync 同步异步模式,async表示异步,sync表示同步,如果设置成异步模式,可以允许生产者以batch形式push数据,这样会极大提高broker性能,推荐设置为异步.
serializer.class kafka.serializer.DefaultEncoder 序列号类,默认序列化类型为byte[]
key.serializer.class Key序列化类,默认同上
partitioner.class kafka.producer.DefaultPartitioner Partition类,默认对key进行hash.
compression.codec none 指定producer消息压缩格式,可选参数为：none / gzip / snappy
compressed.topics null 启用压缩topic名称,若上面参数选择了一个压缩格式,那么压缩仅对本参数指定的topic有效,若本参数为空则对所有topic有效.
message.send.max.retries 3 Producer发送失败时重试次数,若网络出现问题可能会导致不断重试.
queue.buffering.max.ms 5000 启用异步模式时,producer缓存消息时间,比如设置成1000时,它会缓存1秒数据再一次发送出去,这样可以极大增加broker吞吐量,但也会造成时效性降低.
queue.buffering.max.messages 10000 采用异步模式时producer buffer队列里最大缓存消息数量,如果超过这个数值,producer就会阻塞或者丢掉消息.
queue.enqueue.timeout.ms -1 当达到上面参数值时producer阻塞等待时间,如果值设置为0,buffer队列满时producer不会阻塞,消息直接被丢掉,若值设置为-1,producer会被阻塞不会丢消息.
batch.num.messages 200 用异步模式时,一个batch缓存消息数量,达到这个数量值时producer才会发送消息.
send.buffer.bytes 100 * 1024 Socket write buffer size

2.4.3 Consumer 配置信息

属性 默认值 描述
group.id Consumer组ID,相同goup.id的consumer属于同一个组.
zookeeper.connect Consumer的zookeeper连接串,要和broker的配置一致.
consumer.id null 如果不设置会自动生成.
socket.timeout.ms 30 * 1000 网络请求socket超时时间,实际超时时间由max.fetch.wait + socket.timeout.ms 确定.
fetch.message.max.bytes 1024 * 1024 查询topic-partition时允许的最大消息大小,consumer会为每个partition缓存此大小消息到内存,因此这个参数可以控制consumer内存使用量,这个值应该至少比server允许最大消息大小大,以免producer发送消息大于consumer允许消息.
auto.commit.enable true 如果此值设置为true,consumer会周期性把当前消费offset值保存到zookeeper,当consumer失败重启之后将会使用此值作为新开始消费的值.
auto.commit.interval.ms 60 * 1000 Consumer提交offset值到zookeeper周期.
queued.max.message.chunks 2 用来被consumer消费message chunks 数量,每个chunk可以缓存fetch.message.max.bytes大小数据量.
auto.commit.interval.ms 60 * 1000 Consumer提交offset值到zookeeper周期.
queued.max.message.chunks 2 用来被consumer消费message chunks 数量,每个chunk可以缓存fetch.message.max.bytes大小数据量.
consumer.timeout.ms -1 若在指定时间内没有消息消费,consumer将会抛出异常.

3. Kafka 工作流分析

enter image description here

3.1 Kafka 生产过程

3.1.1 写入方式

producer采用推(push)模式将消息发布到broker,每条消息都被追加(append)到分区(patition)中,属于顺序写磁盘(顺序写磁盘效率比随机写内存要高,保障kafka吞吐率).

3.1.2 分区 (Partition)

消息发送时都被发送到一个topic,其本质就是一个目录,而topic是由一些Partition Logs(分区日志)组成,其组织结构如下图所示

enter image description here

每个Partition中的消息都是有序,生产消息被不断追加到Partition log上,其中每一个消息都被赋予了一个唯一的offset值.

分区原因
1.方便在集群中扩展,每个Partition可以通过调整以适应它所在机器,而一个topic又可以有多个Partition组成,因此整个集群就可以适应任意大小数据了.

2.因为以Partition为单位读写,所以可以提高并发.

分区规则
指定patition,则直接使用.
未指定patition但指定key,通过对key的value进行hash出一个patition.
patition和key都未指定,使用轮询选出一个patition.

DefaultPartitioner类

    public int partition(String topic, Object key, byte[] keyBytes,

        Object value, byte[] valueBytes, Cluster cluster) {

        List<PartitionInfo> partitions = cluster.partitionsForTopic(topic);

        intnumPartitions = partitions.size();



        if (keyBytes == null) {

            intnextValue = nextValue(topic);



            List<PartitionInfo> availablePartitions = cluster.availablePartitionsForTopic(topic);



            if (availablePartitions.size() > 0) {

                intpart = Utils.toPositive(nextValue) % availablePartitions.size();



                return availablePartitions.get(part).partition();

            } else {

                // no partitions are available, give a non-available partition

                return Utils.toPositive(nextValue) % numPartitions;

            }

        } else {

            // hash the keyBytes to choose a partition

            return Utils.toPositive(Utils.murmur2(keyBytes)) % numPartitions;

        }

    }

3.1.3 副本 (Replication)

同一个partition可能会有多个replication(对应server.properties配置中的default.replication.factor=N)

没有replication情况下,一旦broker宕机,其上所有patition数据都不可被消费.

同时producer也不能再将数据存于其上的patition,引入replication之后,同一个partition可能会有多个replication,而这时需要在这些replication之间选出一个leader,producer和consumer只与这个leader交互,其它replication作为follower从leader中复制数据.

3.1.4 写入流程

enter image description here

1.producer先从zookeeper /brokers/.../state节点找到该partition的leader.

2.producer将消息发送给该leader.

3.leader将消息写入本地log.

4.followers从leader pull消息,写入本地log后向leader发送ACK.

5.leader收到所有ISR中的replication的ACK后,增加HW(high watermark,最后commit 的offset)向producer发送ACK

3.2 Broker 存储过程

3.2.1 存储方式

物理上把topic分成一个或多个patition(对应server.properties中num.partitions=3配置),每个patition物理上对应一个文件夹(该文件夹存储该patition的所有消息和索引文件).

[root@systemhub711 logss]# ll

total 88

-rw-r--r-- 1 root root    4 Apr 17 22:37 cleaner-offset-checkpoint

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-11

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-14

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-17

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-2

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-20

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-23

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-26

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-29

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-32

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-35

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-38

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-41

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-44

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-47

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-5

drwxr-xr-x 2 root root 4096 Apr 20 15:23 __consumer_offsets-8

-rw-r--r-- 1 root root    4 Apr 20 18:59 log-start-offset-checkpoint

-rw-r--r-- 1 root root   54 Apr 17 14:16 meta.properties

-rw-r--r-- 1 root root  399 Apr 20 18:59 recovery-point-offset-checkpoint

-rw-r--r-- 1 root root  399 Apr 20 19:00 replication-offset-checkpoint

drwxr-xr-x 2 root root 4096 Apr 20 15:28 topic001-0

[root@systemhub711 logss]# cd topic001-0/

[root@systemhub711 topic001-0]# ll

total 8

-rw-r--r-- 1 root root 10485760 Apr 20 15:23 00000000000000000000.index

-rw-r--r-- 1 root root       80 Apr 20 15:28 00000000000000000000.log

-rw-r--r-- 1 root root 10485756 Apr 20 15:23 00000000000000000000.timeindex

-rw-r--r-- 1 root root        8 Apr 20 15:28 leader-epoch-checkpoint

[root@systemhub711 topic001-0]#

3.2.2 存储策略

无论消息是否被消费,kafka都会保留所有消息,有两种策略可以删除旧数据:
1.基于时间 : log.retention.hours=168
2.基于大小 : log.retention.bytes=1073741824
需要注意的是,因为Kafka读取特定消息的时间复杂度为O(1),即与文件大小无关,所以这里删除过期文件与提高Kafka性能无关.

3.2.3 Zookeeper 存储结构

注意：producer不在zk中注册,而消费者在zk中注册.

enter image description here

3.3 Kafka 消费过程

Kafka提供两套Consumer API : 高级Consumer API 和 低级 API

3.3.1 高级 API

高级API 优点

高级API 缺点

3.3.2 低级 API

低级API 优点

低级API 缺点

3.3.3 消费者组

enter image description here

消费者是以consumer group消费者组的方式工作,由一个或者多个消费者组成一个组，共同消费一个topic.

每个分区在同一时间只能由group中的一个消费者读取,但是多个group可以同时消费这个partition.

如图所示,一个由三个消费者组成的group,有一个消费者读取主题中两个分区,另外两个分别读取一个分区,某个消费者读取某个分区,也可以叫做某个消费者是某个分区的拥有者.

在这种情况下,消费者可以通过水平扩展方式同时读取大量消息,另外,如果一个消费者失败了,那么其他group成员会自动负载均衡读取之前失败消费者读取分区.

3.3.4 消费方式

consumer采用pull(拉)模式从broker中读取数据.

push(推)模式很难适应消费速率不同消费者,因为消息发送速率是由broker决定.
它目标是尽可能以最快速度传递消息,但是这样很容易造成consumer来不及处理消息,典型表现就是拒绝服务以及网络拥塞,而pull模式则可以根据consumer消费能力以适当速率消费消息.

对于Kafka而言,pull模式更合适,它可简化broker的设计,consumer可自主控制消费消息速率,同时consumer可以自己控制消费方式——即可批量消费也可逐条消费,同时还能选择不同提交方式从而实现不同传输语义.

pull模式不足之处是,如果kafka没有数据,消费者可能会陷入循环中,一直等待数据到达.

为了避免这种情况,在拉请求中有参数,允许消费者请求在等待数据到达“长轮询”中进行阻塞(并且可选地等待到给定的字节数,以确保大的传输大小)

3.3.5 消费者组案例

测试同一个消费者组中的消费者,同一时刻只能有一个消费者消费.

在systemhub511、systemhub611服务器中修改/opt/module/kafka/config/consumer.properties配置文件中的group.id属性为任意组名.

[root@systemhub511 module]# cd kafka/

[root@systemhub511 kafka]# vim config/consumer.properties

#consumer group id

group.id=systemhub611

[root@systemhub611 module]# cd kafka/

[root@systemhub611 kafka]# vim config/consumer.properties

#consumer group id

group.id=systemhub611

分别启动消费者

[root@systemhub511 kafka]# bin/kafka-console-consumer.sh --zookeeper systemhub511:2181 --topic topic001 --consumer.config config/consumer.properties

Using the ConsoleConsumer with old consumer is deprecated and will be removed in a future major release. Consider using the new consumer by passing [bootstrap-server] instead of [zookeeper].

[067] WARN [systemhub611_systemhub611-1555744110617-1d4d406d], no brokers found when trying to rebalance. (kafka.consumer.ZookeeperConsumerConnector)

[root@systemhub611 kafka]# bin/kafka-console-consumer.sh --zookeeper systemhub511:2181 --topic topic001 --consumer.config config/consumer.properties

Using the ConsoleConsumer with old consumer is deprecated and will be removed in a future major release. Consider using the new consumer by passing [bootstrap-server] instead of [zookeeper].

在systemhub711服务器 启动生产者

[root@systemhub711 kafka]# bin/kafka-console-producer.sh --broker-list systemhub511:9092 --topic topic001

>hello kafka!

查看systemhub511和systemhub611 接收者
同一时刻只有一个消费者接收到消息.

[root@systemhub511 kafka]# bin/kafka-console-consumer.sh --zookeeper systemhub511:2181 --topic topic001 --consumer.config config/consumer.properties

Using the ConsoleConsumer with old consumer is deprecated and will be removed in a future major release. Consider using the new consumer by passing [bootstrap-server] instead of [zookeeper].

[067] WARN [systemhub611_systemhub611-1555744110617-1d4d406d], no brokers found when trying to rebalance. (kafka.consumer.ZookeeperConsumerConnector)

[root@systemhub611 kafka]# bin/kafka-console-consumer.sh --zookeeper systemhub511:2181 --topic topic001 --consumer.config config/consumer.properties

Using the ConsoleConsumer with old consumer is deprecated and will be removed in a future major release. Consider using the new consumer by passing [bootstrap-server] instead of [zookeeper].

hello kafka!

4. Kafka API

4.1 环境准备

<?xml version="1.0" encoding="UTF-8"?>

<project xmlns="http://maven.apache.org/POM/4.0.0"

         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">

    <modelVersion>4.0.0</modelVersion>

    <groupId>com.geekparkhub.core.kafka</groupId>

    <artifactId>kafka_server</artifactId>

    <version>1.0-SNAPSHOT</version>

    <dependencies>

        <dependency>

            <groupId>junit</groupId>

            <artifactId>junit</artifactId>

            <version>3.8.1</version>

            <scope>test</scope>

        </dependency>

        <dependency>

            <groupId>org.apache.kafka</groupId>

            <artifactId>kafka-clients</artifactId>

            <version>0.11.0.0</version>

        </dependency>

        <dependency>

            <groupId>org.apache.kafka</groupId>

            <artifactId>kafka_2.11</artifactId>

            <version>0.11.0.0</version>

        </dependency>

    </dependencies>

</project>

4.2 Kafka 生产者 Java API

4.2.1 Create (过时API)生产者

Create OldProduce.class

package com.geekparkhub.core.kafka.producer;



import kafka.javaapi.producer.Producer;

import kafka.producer.KeyedMessage;

import kafka.producer.ProducerConfig;



import java.util.Properties;



/**

 * Geek International Park | 极客国际公园

 * GeekParkHub | 极客实验室

 * Website | https://www.geekparkhub.com/

 * Description | Open开放 · Creation创想 | OpenSource开放成就梦想 GeekParkHub共建前所未见

 * HackerParkHub | 黑客公园枢纽

 * Website | https://www.hackerparkhub.com/

 * Description | 以无所畏惧的探索精神 开创未知技术与对技术的崇拜

 * GeekDeveloper : JEEP-711

 *

 * @author system

 * <p>

 * OldProduce

 * <p>

 */



public class OldProduce {

    @SuppressWarnings("deprecation")

    public static void main(String[] args) {

        /**

         * Configuration information

         * 配置信息

         */

        Properties props = new Properties();



        /**

         * Kafka configuration information

         * Kafka配置信息

         */

        props.put("metadata.broker.list", "systemhub511:9092");



        /**

         * Response level

         * 应答级别

         */

        props.put("request.required.acks", "1");



        /**

         * K value serialization

         * K值 序列化

         */

        props.put("serializer.class", "kafka.serializer.StringEncoder");



        /**

         * Instantiate producer object

         * 实例化 生产者对象

         */

        Producer<Integer, String> producer = new Producer<Integer, String>(new ProducerConfig(props));



        /**

         * Send data

         * 发送数据

         */

        KeyedMessage<Integer, String> message = new KeyedMessage<Integer, String>("topic001", "Hello,World");

        producer.send(message);



    }

}

4.2.2 Create (新API)生产者

Create NewProduce.class

package com.geekparkhub.core.kafka.producer;



import org.apache.kafka.clients.producer.KafkaProducer;

import org.apache.kafka.clients.producer.Producer;

import org.apache.kafka.clients.producer.ProducerRecord;

import java.util.Properties;



/**

 * Geek International Park | 极客国际公园

 * GeekParkHub | 极客实验室

 * Website | https://www.geekparkhub.com/

 * Description | Open开放 · Creation创想 | OpenSource开放成就梦想 GeekParkHub共建前所未见

 * HackerParkHub | 黑客公园枢纽

 * Website | https://www.hackerparkhub.com/

 * Description | 以无所畏惧的探索精神 开创未知技术与对技术的崇拜

 * GeekDeveloper : JEEP-711

 *

 * @author system

 * <p>

 * NewProduce

 * <p>

 */



public class NewProduce {

    public static void main(String[] args) {



        /**

         * Configuration information

         * 配置信息

         */

        Properties props = new Properties();



        /**

         * Kafka configuration information

         * Kafka配置信息

         */

        props.put("bootstrap.servers", "systemhub511:9092");



        /**

         * Response level

         * 应答级别

         */

        props.put("acks", "all");



        /**

         * number of retries

         * 重试次数

         */

        props.put("retries", 0);



        /**

         * Cache size

         * 缓存大小

         */

        props.put("batch.size", 16384);



        /**

         * Submission delay

         * 提交延时

         */

        props.put("linger.ms", 1);



        /**

         * Cache mode

         * 缓存方式

         */

        props.put("buffer.memory", 33554432);



        /**

         * K value serialization

         * K值 序列化

         */

        props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");



        /**

         * V value serialization

         * V值 序列化

         */

        props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");



        /**

         * Instantiate producer object

         * 实例化 生产者对象

         */

        KafkaProducer<String, String> producer = new KafkaProducer<>(props);



        /**

         * Cycling data

         * 循环发送数据

         */

        for (int i = 0; i < 10; i++) {

            producer.send(new ProducerRecord<String, String>("topic001", String.valueOf(i)));

            System.out.println("Result is = " + producer.toString());

        }



        /**

         * Close resource

         * 关闭资源

         */

        producer.close();



    }

}

[root@systemhub511 kafka]# bin/kafka-console-consumer.sh --zookeeper systemhub511:2181 --topic topic001

Using the ConsoleConsumer with old consumer is deprecated and will be removed in a future major release. Consider using the new consumer by passing [bootstrap-server] instead of [zookeeper].

0

2

4

6

8

1

3

5

7

9

4.2.3 Create (新API)生产者回调函数

Create CallBackProduce.class

package com.geekparkhub.core.kafka.producer;



import org.apache.kafka.clients.producer.Callback;

import org.apache.kafka.clients.producer.KafkaProducer;

import org.apache.kafka.clients.producer.ProducerRecord;

import org.apache.kafka.clients.producer.RecordMetadata;

import java.util.Properties;



/**

 * Geek International Park | 极客国际公园

 * GeekParkHub | 极客实验室

 * Website | https://www.geekparkhub.com/

 * Description | Open开放 · Creation创想 | OpenSource开放成就梦想 GeekParkHub共建前所未见

 * HackerParkHub | 黑客公园枢纽

 * Website | https://www.hackerparkhub.com/

 * Description | 以无所畏惧的探索精神 开创未知技术与对技术的崇拜

 * GeekDeveloper : JEEP-711

 *

 * @author system

 * <p>

 * CallBackProduce

 * <p>

 */



public class CallBackProduce {

    public static void main(String[] args) {

        /**

         * Configuration information

         * 配置信息

         */

        Properties props = new Properties();



        /**

         * Kafka configuration information

         * Kafka配置信息

         */

        props.put("bootstrap.servers", "systemhub511:9092");



        /**

         * Response level

         * 应答级别

         */

        props.put("acks", "all");



        /**

         * number of retries

         * 重试次数

         */

        props.put("retries", 0);



        /**

         * Cache size

         * 缓存大小

         */

        props.put("batch.size", 16384);



        /**

         * Submission delay

         * 提交延时

         */

        props.put("linger.ms", 1);



        /**

         * Cache mode

         * 缓存方式

         */

        props.put("buffer.memory", 33554432);



        /**

         * K value serialization

         * K值 序列化

         */

        props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");



        /**

         * V value serialization

         * V值 序列化

         */

        props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");



        /**

         * Instantiate producer object

         * 实例化 生产者对象

         */

        KafkaProducer<String, String> producer = new KafkaProducer<>(props);



        /**

         * Cycling data

         * 循环发送数据

         */

        for (int i = 0; i < 10; i++) {

            producer.send(new ProducerRecord<String, String>("topic001", String.valueOf(i)), new Callback() {

                @Override

                public void onCompletion(RecordMetadata metadata, Exception exception) {

                    /**

                     * Determine if the exception is empty

                     * 判断exception是否为空

                     */

                    if (exception == null) {

                        System.out.println("Data Sent Successfully !");

                        /**

                         * Get metadata information : offset & partition

                         * 获取元数据信息 : offset & partition

                         */

                        long offset = metadata.offset();

                        int partition = metadata.partition();

                        System.out.println("Offset is = " + offset + " -- & -- Partition is = " + partition);

                    } else {

                        System.out.println("Data Transmission Failed !");

                    }

                }

            });

        }



        /**

         * Close resource

         * 关闭资源

         */

        producer.close();

    }

}

enter image description here

[root@systemhub511 kafka]# bin/kafka-console-consumer.sh --zookeeper systemhub511:2181 --topic topic001

Using the ConsoleConsumer with old consumer is deprecated and will be removed in a future major release. Consider using the new consumer by passing [bootstrap-server] instead of [zookeeper].

0

2

4

6

8

1

3

5

7

9

4.2.4 自定义分区生产者

package com.geekparkhub.core.kafka.producer;



import org.apache.kafka.clients.producer.Partitioner;

import org.apache.kafka.common.Cluster;



import java.util.Map;



/**

 * Geek International Park | 极客国际公园

 * GeekParkHub | 极客实验室

 * Website | https://www.geekparkhub.com/

 * Description | Open开放 · Creation创想 | OpenSource开放成就梦想 GeekParkHub共建前所未见

 * HackerParkHub | 黑客公园枢纽

 * Website | https://www.hackerparkhub.com/

 * Description | 以无所畏惧的探索精神 开创未知技术与对技术的崇拜

 * GeekDeveloper : JEEP-711

 *

 * @author system

 * <p>

 * CustomPartitioner

 * <p>

 */



public class CustomPartitioner implements Partitioner {



    /**

     * Empty reference constructor

     * 空参构造器

     */

    public CustomPartitioner() {

        super();

    }



    /**

     * 复写分区方法

     *

     * @param topic

     * @param key

     * @param keyBytes

     * @param value

     * @param valueBytes

     * @param cluster

     * @return

     */

    @Override

    public int partition(String topic, Object key, byte[] keyBytes, Object value, byte[] valueBytes, Cluster cluster) {

        return 0;

    }



    /**

     * Close resource

     * 关闭资源

     */

    @Override

    public void close() {

    }



    /**

     * Configuration information

     * 配置信息

     *

     * @param configs

     */

    @Override

    public void configure(Map<String, ?> configs) {

    }

}



package com.geekparkhub.core.kafka.producer;



import org.apache.kafka.clients.producer.Callback;

import org.apache.kafka.clients.producer.KafkaProducer;

import org.apache.kafka.clients.producer.ProducerRecord;

import org.apache.kafka.clients.producer.RecordMetadata;

import java.util.Properties;



/**

 * Geek International Park | 极客国际公园

 * GeekParkHub | 极客实验室

 * Website | https://www.geekparkhub.com/

 * Description | Open开放 · Creation创想 | OpenSource开放成就梦想 GeekParkHub共建前所未见

 * HackerParkHub | 黑客公园枢纽

 * Website | https://www.hackerparkhub.com/

 * Description | 以无所畏惧的探索精神 开创未知技术与对技术的崇拜

 * GeekDeveloper : JEEP-711

 *

 * @author system

 * <p>

 * CustomProducePartitioner

 * <p>

 */



public class CustomProducePartitioner {

    public static void main(String[] args) {

        /**

         * Configuration information

         * 配置信息

         */

        Properties props = new Properties();



        /**

         * Kafka configuration information

         * Kafka配置信息

         */

        props.put("bootstrap.servers", "systemhub511:9092");



        /**

         * Response level

         * 应答级别

         */

        props.put("acks", "all");



        /**

         * number of retries

         * 重试次数

         */

        props.put("retries", 0);



        /**

         * Cache size

         * 缓存大小

         */

        props.put("batch.size", 16384);



        /**

         * Submission delay

         * 提交延时

         */

        props.put("linger.ms", 1);



        /**

         * Cache mode

         * 缓存方式

         */

        props.put("buffer.memory", 33554432);



        /**

         * K value serialization

         * K值 序列化

         */

        props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");



        /**

         * V value serialization

         * V值 序列化

         */

        props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");



        /**

         * Custom Partition

         * 自定义分区

         */

        props.put("partitioner.class", "com.geekparkhub.core.kafka.producer.CustomPartitioner");



        /**

         * Instantiate producer object

         * 实例化 生产者对象

         */

        KafkaProducer<String, String> producer = new KafkaProducer<>(props);



        /**

         * Cycling data

         * 循环发送数据

         */

        for (int i = 0; i < 10; i++) {

            producer.send(new ProducerRecord<String, String>("topic001", String.valueOf(i)), new Callback() {

                @Override

                public void onCompletion(RecordMetadata metadata, Exception exception) {

                    /**

                     * Determine if the exception is empty

                     * 判断exception是否为空

                     */

                    if (exception == null) {

                        System.out.println("Data Sent Successfully !");

                        /**

                         * Get metadata information : offset & partition

                         * 获取元数据信息 : offset & partition

                         */

                        long offset = metadata.offset();

                        int partition = metadata.partition();

                        System.out.println("Offset is = " + offset + " -- & -- Partition is = " + partition);

                    } else {

                        System.out.println("Data Transmission Failed !");

                    }

                }

            });

        }



        /**

         * Close resource

         * 关闭资源

         */

        producer.close();

    }

}

enter image description here

[root@systemhub511 kafka]# bin/kafka-console-consumer.sh --zookeeper systemhub511:2181 --topic topic001

Using the ConsoleConsumer with old consumer is deprecated and will be removed in a future major release. Consider using the new consumer by passing [bootstrap-server] instead of [zookeeper].

0

1

2

3

4

5

6

7

8

9

4.3 Kafka 消费者 Java API

4.3.1 Create (过时API)消费者

package com.geekparkhub.core.kafka.consumer;



import kafka.consumer.Consumer;

import kafka.consumer.ConsumerConfig;

import kafka.consumer.ConsumerIterator;

import kafka.consumer.KafkaStream;

import kafka.javaapi.consumer.ConsumerConnector;

import java.util.HashMap;

import java.util.List;

import java.util.Map;

import java.util.Properties;



/**

 * Geek International Park | 极客国际公园

 * GeekParkHub | 极客实验室

 * Website | https://www.geekparkhub.com/

 * Description | Open开放 · Creation创想 | OpenSource开放成就梦想 GeekParkHub共建前所未见

 * HackerParkHub | 黑客公园枢纽

 * Website | https://www.hackerparkhub.com/

 * Description | 以无所畏惧的探索精神 开创未知技术与对技术的崇拜

 * GeekDeveloper : JEEP-711

 *

 * @author system

 * <p>

 * OldCustomConsumer

 * <p>

 */



public class OldCustomConsumer {



    @SuppressWarnings("deprecation")

    public static void main(String[] args) {

        Properties properties = new Properties();

        properties.put("zookeeper.connect", "systemhub511:2181");

        properties.put("group.id", "test");

        properties.put("zookeeper.session.timeout.ms", "500");

        properties.put("zookeeper.sync.time.ms", "250");

        properties.put("auto.commit.interval.ms", "1000");



        ConsumerConnector consumer = Consumer.createJavaConsumerConnector(new ConsumerConfig(properties));

        HashMap<String, Integer> topicCount = new HashMap<>();

        topicCount.put("topic001", 1);

        Map<String, List<KafkaStream<byte[], byte[]>>> consumerMap = consumer.createMessageStreams(topicCount);

        KafkaStream<byte[], byte[]> stream = consumerMap.get("topic001").get(0);

        ConsumerIterator<byte[], byte[]> it = stream.iterator();

        while (it.hasNext()) {

            System.out.println(new String(it.next().message()));

        }

    }

}

4.3.2 Create (新API)消费者

package com.geekparkhub.core.kafka.consumer;



import org.apache.kafka.clients.consumer.ConsumerRecord;

import org.apache.kafka.clients.consumer.ConsumerRecords;

import org.apache.kafka.clients.consumer.KafkaConsumer;



import java.util.Arrays;

import java.util.Properties;



/**

 * Geek International Park | 极客国际公园

 * GeekParkHub | 极客实验室

 * Website | https://www.geekparkhub.com/

 * Description | Open开放 · Creation创想 | OpenSource开放成就梦想 GeekParkHub共建前所未见

 * HackerParkHub | 黑客公园枢纽

 * Website | https://www.hackerparkhub.com/

 * Description | 以无所畏惧的探索精神 开创未知技术与对技术的崇拜

 * GeekDeveloper : JEEP-711

 *

 * @author system

 * <p>

 * CustomConsumer

 * <p>

 */



public class CustomConsumer {



    public static void main(String[] args) {



        /**

         * Configuration information

         * 配置信息

         */

        Properties props = new Properties();



        /**

         * Kafka configuration information

         * Kafka配置信息

         */

        props.put("bootstrap.servers", "systemhub511:9092");



        /**

         * Consumer group ID

         * 消费者组ID

         */

        props.put("group.id", "test");



        /**

         * Set auto-submit offset

         * 设置自动提交offset

         */

        props.put("enable.auto.commit", "true");



        /**

         * Submission delay

         * 提交延时

         */

        props.put("auto.commit.interval.ms", "1000");



        /**

         * K value Deserialization

         * K值 反序列化

         */

        props.put("key.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");



        /**

         * V value Deserialization

         * V值 反序列化

         */

        props.put("value.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");



        /**

         * Instantiation Consumer Object

         * 实例化 消费者对象

         */

        KafkaConsumer<String, String> consumer = new KafkaConsumer<>(props);



        /**

         * Specify Topic

         * 指定Topic

         */

        consumer.subscribe(Arrays.asList("topic001", "topic002", "topic003"));



        while (true) {

            /**

             * retrieve data

             * 获取数据

             */

            ConsumerRecords<String, String> consumerRecords = consumer.poll(100);



            /**

             * Loop output

             * 循环输出

             */

            for (ConsumerRecord<String, String> records : consumerRecords) {

                String topic = records.topic();

                int partition = records.partition();

                String value = records.value();

                System.out.println("Topic is = " + topic + " -- & -- Partition is = " + partition + " -- & -- Value is = " + value);

            }

        }

    }

}

[root@systemhub711 kafka]# bin/kafka-console-producer.sh --broker-list systemhub511:9092 --topic topic001

[root@systemhub711 kafka]# bin/kafka-console-producer.sh --broker-list systemhub511:9092 --topic topic002

root@systemhub711 kafka]# bin/kafka-console-producer.sh --broker-list systemhub511:9092 --topic topic003

[root@systemhub711 kafka]# bin/kafka-console-producer.sh --broker-list systemhub511:9092 --topic topic001

>hello topic001

[root@systemhub711 kafka]# bin/kafka-console-producer.sh --broker-list systemhub511:9092 --topic topic002

>hello topic002

[root@systemhub711 kafka]# bin/kafka-console-producer.sh --broker-list systemhub511:9092 --topic topic003

>hello topic003

enter image description here

5. Kafka Producer拦截器

6. Kafka Streams

7. 修仙之道 技术架构迭代 登峰造极之势

Alt text

💡如何对该开源文档进行贡献💡

  1. Blog内容大多是手敲,所以难免会有笔误,你可以帮我找错别字。

  2. 很多知识点我可能没有涉及到,所以你可以对其他知识点进行补充。

  3. 现有的知识点难免存在不完善或者错误,所以你可以对已有知识点的修改/补充。

  4. 💡欢迎贡献各领域开源野生Blog&笔记&文章&片段&分享&创想&OpenSource Project&Code&Code Review

  5. 🙈🙈🙈🙈🙈🙈🙈🙈🙈🙈🙈 issues: geekparkhub.github.io/issues 🙈🙈🙈🙈🙈🙈🙈🙈🙈🙈🙈

希望每一篇文章都能够对读者们提供帮助与提升,这乃是每一位笔者的初衷

💌感谢您的阅读 欢迎您的留言与建议💌

捐助 项目的发展离不开你的支持,请开发者喝杯☕Coffee☕吧!

enter image description here

致谢：

捐助时请备注 UserName

ID UserName Donation Money Consume
1 Object WeChatPay 5RMB 一杯可乐
2 泰迪熊看月亮 AliPay 20RMB 一杯咖啡
3 修仙道长 WeChatPay 10RMB 两杯可乐

License 开源协议

Apache License Version 2.0